The mood stabilizers lithium and valproate are both effective in the treatment of bipolar disorder; however, their therapeutic mechanisms remain unclear. Because of the delayed onset of clinical efficacy (days to weeks), it has been proposed that adaptive changes in gene expression, rather than initial pharmacological actions, may be directly responsible. Using cDNA microarray as the initial screening method, we discovered that chronic administration of both agents at therapeutic doses increased the expression of BAG-1 (bcl-2 associated athanogene) in rat hippocampus. Furthermore, these findings were validated in the hippocampus at the protein level, the effects were seen in a time frame consistent with therapeutic effects, and were specific for mood stabilizers. BAG -1 is an important chaperone of bcl-2 (B-cell CLL/lymphoma 2), and enhances bcl-2_s anti-apoptotic functions; furthermore, through interaction with raf (v-raf-1 murine leukemia viral oncogene homolog 1), BAG-1 is able to activate ERK (extracellular signal-regulated protein kinase) MAP (mitogen-activated protein) kinases. Consistent with this, we previously found that lithium and valproate activate ERK MAP kinases. Bag-1 also inhibits GR (glucocorticoid receptor) activation, which may counteract the deleterious effects of hypercortisolemia seen in bipolar disorder. Anti-GR antibody immunostaining plus double staining with DAPI (4',6-Diamidino-2-phenylindole) showed either lithium or VPA, at therapeutically relevant levels, inhibited dexamethasone induced GR nuclear translocation. In addition, glucocorticoid response element (GRE) transfection assay showed lithium, at therapeutically relevant levels, inhibited GR activity in cultured human cells. Evaluated through siRNA (short interference RNA) silencing of BAG-1, the inhibition of mood stabilizers to GR nuclear translocation and to GR activity is mediated, at least in part, by BAG-1. The effect that BAG-1 inhibits glucocorticoid activation suggests mood stabilizers may counteract the deleterious effects of hypercortisolemia seen in bipolar disorder by up-regulating BAG-1. The role of BAG-1 in behavioral plasticity relevant to mood disorders was further investigated in wild-type and neuron-selective BAG-1 transgenic mice using a battery of behavioral tests. The BAG1 mice appeared normal in growth and in neurological and sensory tests. Mice were then subjected to behavioral tests related to anxiety, depression, and mania. BAG1 transgenic mice showed less anxious-like behavior on the elevated plus maze test. The mice did not differ from wild-type mice in the forced swim test, and exhibited comparable rates of developing helplessness behavior in the learned helplessness paradigm. However, the BAG1 transgenic mice showed higher spontaneous recovery rates from the helplessness behavior. On mania-related tests, BAG1 transgenic mice recovered much faster in the amphetamine-induced hyperlocomotion test, and displayed a clear resistance to cocaine-induced behavioral sensitization. BAG1 transgenic mice exhibited specific hippocampal neurochemical alterations including increased Hsp70 and decreased FKBP51 levels. Together, the data support that BAG-1 represent a novel, highly therapeutically relevant target in the long-term treatment of bipolar disorder and play role in mood stability. This role of BAG-1 will be further investigated in the BAG-1 knockout mice.